wayfinder
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads and processes external content from issue trackers, which creates a potential surface for indirect prompt injection.
- Ingestion points: The skill instructs the agent to read issue bodies, discussion threads, and planning context from external GitHub or GitLab repositories.
- Boundary markers: There are no specific delimiters or instructions to ignore embedded commands within the external content provided in the skill instructions.
- Capability inventory: The agent is authorized to write to trackers, manage issue states, and generate executable prototypes based on its analysis.
- Sanitization: While the agent is cautioned not to expose secrets, the skill does not provide technical sanitization mechanisms to prevent external input from overriding system instructions.
- [DYNAMIC_EXECUTION]: The skill incorporates a prototyping phase that involves creating "runnable experiments." This requires the agent to generate and execute code at runtime based on requirements that may be influenced by untrusted external inputs from research and discussions.
Audit Metadata